Sir Nigel Gresley - The L.N.E.R.’s First C.M.E.

I agree with most of those points, but think the question of where the axes cross is less clearcut.

Taking a zero point for the mileage and/or availability is self evidently statistically correct, but it's less clear whether that is actually the most useful interpretation.

The advantage of bunching up like this is that it draws out the level of change. So, the decline from 1943 to 1944 in both availability (from 75% to 73%, which is 2.7% overall) and mileage (43k to 41k, which is a 4.7% decline) is much more apparent in the first graph. Bringing my 21st century managerialist perspective to this data, I would have expected interest to be less in those absolute levels of performance, and more in the level and direction of change to first stabilise and then improve fleet performance.

 

Yes, the Ops Manager in me though is saying to his boss "look how consistent our performance is year on year ..."

@Jimc has put the two sides of the argument very clearly above; as do the two graphs of the same data @S.A.C. Martin showed. The initial graph makes it look like the operational performance is all over the place; the second makes it look like nothing to worry about - yet it is the same underlying data. So an important point is to try to find out what level of variation was considered significant. I suspect a 10% change in availability between two things (*) was probably an issue worth following up; and a 1% variation was lost in the noise. But that's just a guess.

I'll dig my heals in though about bar charts for mileage data and a line for availability ...

(*) i.e. 10% variation between years; or classes; or sheds - i.e. the level of variation when comparing like with like

Tom

 

Thanks Fred , the decrease i understand and the fall off in mileage corresponds, but how do explain the increase... is it because scheduled overhauls and maintenance was being 'deferred' creating more availability... until the inevitable breakdowns ?

 

Here's an attempt at that Tom:



I will sort out the decimal places in due course.

Thanks all for your thoughts, I will respond properly after work today.

 

Completely agree, and I've no view of materiality. If I were doing this now on this dataset, I'd also be looking quite closely at this data on a more granular basis - by week or month, and shed vs shed. @S.A.C. Martin and his would be readers will be glad that I don't propose that he drill down to that level of detail, even if it might highlight some interesting "what ifs".

 

Excel does lots of stuff with number formats - I can lose hours happily fiddling...

 

Mind you its also true that small changes in large numbers can have a big effect - especially if its the difference between marginally loss making and marginally proitable. That's the trouble with trying to establish the real facts - its so much more difficult than moderately informed speculation!

 

I did consider it, but there's so much data that it would take an absolute age to root through it all. The Use of Engine Power document, incidentally, splits the statistics as below:

 

My immediate comment might be to start by thinking what you wish to demonstrate with the data, and perhaps more importantly how do these figures compare with a similar period prior to WW2?

 

It wasn’t an entirely serious suggestion! The reality is that the dataset is large and complex enough that an in depth investigation would require some serious statistical analysis, and you probably will never get some of the data you’d need to reach a full understanding.

As others have asked, much depends on what question you’re trying to answer


Sent from my iPhone using Tapatalk

 

Colours are spot on !

 

Excellent! A livery debate about spreadsheets! Now, is that the Darlington or Doncaster green?

Tom

 

Doncaster.

Not being entirely serious, but I did colour code this spreadsheet to LNER colours...!

 

Whilst I suspect that more repairs were undertaken at main depots rather than at workshops I do not have the record cards to confirm that assumption; this may be something that Simon's analysis could highlight. Even if the assumption proves correct it will not identify (a) how much work was undertaken and (b) was it just sufficient to keep the locomotive operating or a full service to minimise need to visit a workshops facility hence that swill depend on contemporary reports / sources of the time (e.g. Railway Observer).

 

Simon has already made very good use of the 1940s statistics in his book on Edward Thompson. It will be most interesting if mileage and availability figures come to light for earlier years, allowing analysis to be extended further back in time.

The LNER loco fleet stayed almost the same size through the 1920s - totals of 7392 at Grouping and 7378 at the end of 1929. But the total then falls by over 12% through the 1930s to a total of 6476 by end of 1939. Some of this drop was undoubtedly linked to the 1930s trade recession. But I wonder whether the LNER was also achieving higher utilization from its loco fleet, which should be reflected in higher average mileages.

Figures taken from RCTS Locos of LNER Part 10A.

 

I am delighted to say that I have been given a copy of a similar document from earlier in the LNERs existence. Watch this space.

Newer locomotive classes generally replaced 1.5 to 2 locomotives in terms of work, roughly, provided availability was good.

 

On this day in 1876, Sir Nigel Gresley was born!
Pictured here in 1938 next to the A4 locomotive No 4498, named after him.

 

I can see that you are lining up some interesting new material for your book on Sir Nigel. I seem to recall plenty of discussion in previous books about the LMS achieving improved loco utilization but not much about the position on the LNER. But as noted in my previous post, the LNER loco fleet declined by about 12% during the 1930s, which may indicate some improved utilization as well as straightforward reductions due to the trade depression.

A couple of questions arise in respect of these withdrawals. The first is what was being withdrawn? A cursory look suggests that the bulk were low-power elderly (pre-1900) engines. But there were also notable early withdrawals of some relatively modern classes, such as the NB Atlantics (C11), NE Worsdell 4-6-0s (B13/14) and GN "Long Tom" 0-8-0s (Q1/2).

Former Hull & Barnsley engines fared poorly under the LNER, with some 90% withdrawn by 1939 and only a few surviving to reach BR. A likely factor here is that H&B engines existed in small numbers when compared with equivalent types from the NE and other major constituents. Locos taken-over in 1937 from the M&GN and the Metropolitan were also soon scrapped. This mirrors the fate of the loco fleets of the smaller LMS constituents. Exceptions to this pattern were the Highland and GNSR, whose engines survived well. In the case of the GNSR, might a factor have been that a single boiler type (Diagram 88) fitted more than half the fleet?

A second question is who on the LNER decided what was to be withdrawn? Did Gresley lay down company-wide policies and directions on this issue? Or did he leave the matter to the Area Mechanical Engineers and Running Superintendents?

 

Yes, this is what I am (privately) hoping it will show, however early indications are that this is not the case actually. There is in my view no such thing as "straightforward reductions" on the LNER, there are always caveats.

I have previously produced a heat map showing length of service and withdrawal dates so I know the answer to this, but there's an interesting historical record to be told in terms of the classes that stayed in existence (probably longer than they should, including one utterly bizarre story of a class withdrawn in the 1930s and - we have discovered - being reintroduced in the 1940s in a limited way).

Locomotive running superintendents had the power to keep or scrap locomotives, provided the financial side of the LNER approved the expenditure/retention/etc. There were company wide policies on how this subject was approached but ultimately it was the superintendents locally who worked on these matters.

 

Confirming that you form your views according to the actual evidence (unlike some).

In that case the early withdrawal of some classes

suggests that perhaps those classes weren't much liked even on their home turf.